Selecting, transforming, visualizing and analyzing data through the Data Explorer

The Data Explorer allows you to select, transform, visualize and analyze your production data in the custom way that best suits you, including in real-time.

Date: November 7, 2019 | Wizata platform version 3.10.7

1. Selecting the data to display

Select data points on the data point selection and choose what data type to display.

2. Selecting the time interval of data

Click on the time interval on the top menu of the Charts element to select a time interval of the data. Or if you prefer use, the 5 default options (Last Hour, Last Day, Last Week, Last 30 Days or last 90 days on the left-hand side). When exploring very granular data, precision is important. In the Data Explorer it’s possible to explore such data by choosing seconds in the From-To range and you can limit your data view to a few seconds.

A visual interactive interface of the timeline of the selected data points is displayed at the top of the screen. Click on a data point timeline to be able to change the start time, stop time, total interval, or to remove the data point from the selection.

Please note, that if you are trying to visualize a period of time extensively long, this exercise might cause unnecessary load on the underlying data structures and an extensive waiting time. For this reason, if data loading in Data Explorer or Chart Widget takes longer than 10 seconds, a button to cancel data loading will appear. Just click this button if you need to stop loading the data for any reason.

3. Selecting the granularity of the displayed data 

The top menu of the Charts element allows you to select the granularity of the data. This way you can adapt it to the selected time interval and see the data in that granularity that shows the phenomenon you are studying. The options are Auto, (as Autodata, Batch, Raw or Custom).

If you select Custom, you will be able to input the granularity yourself by day, hour, minute and second. For example 33 seconds as shown in the image.

4. Transforming how the data is displayed

The 'paintbrush icon' next to data points of the Data Points selection element allows you to change individual display settings of the selected sensors.

You can then pick a 'Type' of visualization, choosing between Line, Line Rough, Line Step, Bar, Scatter, and Histogram (with an additional 'Histogram Type' setting: Square-root, Freedman-Diaconis, Scott's normal reference, and Sturges).

If you want to check the evolution of a feature over time, you can pick Line, Line Rough or Line Step. The difference between between the three types of lines is the interpolation method used to connect the data values.

You can use a Bar chart to see the evolution of a cumulative data point over time.

A Scatter plot will show you the relationship between two data points. You will be able to explore correlations with this type of plot.

A Histogram will give you an insight of the distribution of values for a data point. Also, you will be able to identify outliers with this type of plot.

You can check the Show Values setting to display the values recorded on the chart. It can be useful to display data availability.

When data hasn't been recorded, you can check the Data Interpolation setting to interpolate the missing data between recorded values.

The 'paintbrush icon' next to a group of the Data Points selection element allows you to change individual display settings of the sensors in the group.

The 'paintbrush icon' of the top menu of the Charts element allows you to change global display settings of all selected sensors.

Click on the 'paintbrush icon' of the top menu of the Charts element and then check 'Multigrid' to display data of multiple selected data points on different charts.

Click on the 'paintbrush icon' of the top menu of the Charts element and input a number in the 'Columns' field to display data of multiple selected data points on different charts and in multiple columns. In practice, this settings requires to apply the 'Multigrid' option in parallel.

Click on the 'paintbrush icon' of the top menu of the Charts element and then check 'Normalize' to normalize the data of multiple selected data points.

To analyze how different data points correlate via a scatterplot, drag & drop a selected data point onto another selected data point to create a group, then click on the paintbrush icon of the group and select 'Scatter 2D' under the 'Type' tooltip selection.

You can then pick a 'Regression Type', choosing between Linear, Through the Origin, Exponential, Polynomial, Logarithmic, LOESS (with an additional Bandwidth setting), Power Law, and Quadratic.

After picking the type of Regression that fits better your data, you will obtain the coefficients of this regression fit. These coefficients define the relationship between the two data points plotted captured as an analytical expression. With this analytical expression you are able to predict for each value of the variable on the horizontal axis, the corresponding value for the variable on the vertical axis.

You can perform multiple fixed-length regressions on subsets of plotted data. We called these 'Range Regressions' because they apply to a small range of data. You can calculate the trend based on the last 10 values, for example.

5. Highlighting data based on custom conditions

You can set conditions to highlight specific values on the Charts of the Data Explorer. For that, click on 'Update Condition' button on the bottom left of the Data Explorer.

  • Input a name, input source, output source for the new condition. Select the condition and value that should filter the data to highlight. Select the 'Inverse' toggle if you wish to highlight the data not selected by the condition.
  • Click 'Save' to view the highlighted data.
  • Click on the color button if you wish to change the color that highlights the data for each condition.
  • You can always edit and delete existing conditions by clicking on the icons next to the list of existing conditions.